PROJECT SUMMARY/ABSTRACT Each year 2 million older adults receive antipsychotic medications (APMs) in the hospital, mainly for delirium. Although the harms of APMs are well documented and their use has declined in older adults with dementia, the scope and risk of off-label APM exposure in hospitalized older adults remain largely unknown. The objective of this application is to determine the prescribing trends and risks of APMs and their alternatives in hospitalized older adults with postoperative delirium. The research team will analyze data from 2 national inpatient data- bases and detailed clinical data from 5 NIH-funded cohorts to generate evidence on the comparative safety of APMs and their alternatives that is generalizable to real-world patients. The central hypothesis is that off-label APM use is associated with increased risk of adverse events compared with no use in hospitalized older adults with postoperative delirium; in particular, newer ?atypical? APMs may be more harmful than conventional ?typi- cal? APM for short-term use. The applicant's preliminary data showed that, although atypical APMs may be as harmful as typical APM in hospitalized older surgical patients, use of atypical APMs and certain psychoactive drugs has been rising in recent years. The APM prescribing rates also vary more than 30 folds across the US hospitals, which suggests potentially inappropriate prescribing. Lack of safety data on APMs for hospitalized older adults, as well as little evidence on safer alternatives seems to contribute to these trends. The rationale for this application is that the available safety data from dementia patients may not generalize to hospitalized older surgical patients and that evidence on the comparative safety of APMs and other psychoactive drugs in routine care patients cannot be generated from clinical trials alone. This 3-year proposal has 3 specific aims: 1) determine the risk of in-hospital mortality, institutional discharge, prolonged length of stay, and readmission associated with different APMs in older patients with delirium after major surgery; 2) determine the risk of in- hospital non-fatal adverse events associated with different APMs; and 3) identify psychoactive drugs that may be prescribed as therapeutic alternatives to APMs by examining longitudinal prescribing trends. The investiga- tors will apply innovative research methodologies of leveraging existing large claims datasets and cohort data to enhance generalizability to routine care populations, increase statistical power to address safety for patient subgroups and individual drugs, and reduce bias. By conducting validation of case-identification algorithms, the investigators will improve how in-hospital adverse events are measured in inpatient drug safety research. The impact of this research is significant, because high-quality evidence on the comparative safety of APMs and their alternatives can guide clinicians to reduce excessive APM use and promote rational and safe APM prescribing, thereby improving safety and clinical care of older surgical patients.